AD 6698 Generative AI for Business Analytics

Schedule

1 Course Schedule

Please adhere to the due dates posted on the course website. Late work is not accepted.

Day A1 A2 A3
Day 1 Tuesday Thursday Wednesday
Day 2 Wednesday Friday Thursday
Day 3 Thursday Saturday Friday
Day 4 Friday Sunday Saturday
Day 5 Saturday Monday Sunday
Day 6 Sunday Tuesday Monday
Day 7 Monday Wednesday Tuesday

1.1 Course Schedule

Lecture Title Description Online On-Campus
L0.1 Dev Setup: GitHub, Jupyter, VS Code, Colab Course tools setup and basic version control with GitHub, notebooks, and IDEs. nan nan
L0.2 Introduction to Neural Networks Understanding the basics of neural networks as a foundation for generative models. nan nan
L1.1 Introduction to Course, Natural Language, and Generative AI Landscape Overview of generative AI, its business relevance, and natural language as a data type. nan nan
L1.2 Introduction to Natural Language Processing Core NLP concepts: tokenization, POS tagging, parsing, and text pre-processing. nan nan
L2.1 Prompt Engineering Fundamentals Learn prompt design, zero-/few-shot techniques, and ICL with business use cases. nan nan
L2.2 Tokenization, Embeddings, and Vector Semantics Deep dive into tokenization methods, word embeddings, and semantic search foundations. nan nan
L3.1 LLM Internals: Attention, Transformers, and Positional Encoding Explore transformer architecture, attention mechanisms, and context windows. nan nan
L3.2 Fine-Tuning Pipelines & Data Handling Set up custom model pipelines and format datasets for supervised fine-tuning. nan nan
L4.1 RAG Basics: Embeddings, Chunking, Indexing Understanding retrieval-augmented generation concepts and corpus preparation. nan nan
L4.2 RAG Query Pipelines with LangChain & LlamaIndex Build Q&A systems using vector DBs, LangChain chains, and document loaders. nan nan
L5.1 LoRA, QLoRA, PEFT Techniques Learn parameter-efficient fine-tuning methods to adapt base models. nan nan
L5.2 Embeddings & Similarity Search Work with sentence embeddings for document clustering, search, and relevance. nan nan
L6.1 Memory & Context Management Strategies for working with limited context: caching, summaries, and memory buffers. nan nan
L6.2 Multimodal Models and Use Cases Explore vision-language models and multimodal business applications. nan nan
L7.1 Model Evaluation & GPT-Eval Evaluate generated outputs using automated tools, manual rubrics, and hallucination detection. nan nan
L7.2 Responsible AI & Bias Detection Assess bias, fairness, and ethics in generated outputs using toolkits and guidelines. nan nan
L8.1 Intro to Agents: ReAct, Tool Use Overview of agent-based systems and tool-augmented LLM workflows. nan nan
L8.2 Multi-Agent Systems (AutoGen, CrewAI) Design and orchestrate multi-agent collaborative systems for enterprise tasks. nan nan